Pody for Cursor - MCP Server for Podcast Tools
Pody is the first Israeli podcast platform built for AI-native workflows. We are designing a public Model Context Protocol (MCP) server at https://pody.io/mcp (currently in design, targeting Q3 2026) so that Cursor, Claude, and any other MCP-capable AI assistant can query live studio availability, search podcast catalogs, look up Hebrew podcast terminology, and pull structured industry data without leaving the editor. Connect once; your AI gets real context about the Israeli podcast ecosystem.
Who We Are
Pody (pody.io) is a podcast and video content marketplace operated by MindSeller LTD (ח.פ. 515284891), headquartered in Tel Aviv. Founded in 2024 by Aviv Charuvi, the platform connects three sides of the content market: podcasters who need studios and editing, recording studios that want bookings, and video editors who earn by claiming jobs.
- Studios: 6 verified Israeli studios currently listed, with real-time availability and Podys-based booking.
- Languages: Hebrew, English, Arabic, Russian.
- Storage: Archive-first architecture - every episode is permanently stored on Internet Archive so podcast RSS feeds survive even if Pody goes down. R2 (Cloudflare) serves as a 30-day hot cache.
- Plans: Free / Pro at ₪97/month / Max at ₪197/month (VAT 18% included).
- Identity: Wikidata Q139719199.
API and MCP
The Pody MCP server is currently in design and is not yet live. The planned endpoint is:
https://pody.io/mcp
Once released (Q3 2026), any MCP-compatible client can add it in one line. For Cursor, that means adding a block to your ~/.cursor/mcp.json:
{
"mcpServers": {
"pody": {
"url": "https://pody.io/mcp",
"transport": "http-sse"
}
}
}
Planned tools the server will expose:
| Tool name |
What it returns |
studios.search |
Find recording studios by city, capacity, or available slot. Returns structured JSON with real availability. |
podcasts.search |
Full-text search across public podcast catalog on Pody. Useful for discovery, guest research, or competitive context. |
glossary.lookup |
Authoritative definitions for Hebrew podcast and audio production terminology. |
state-of-industry.query |
Aggregated, anonymized statistics about the Israeli podcast industry (episode counts, genre breakdown, publishing cadence). |
All tools return JSON. No authentication is required for read-only studio search and glossary lookups. Write operations (booking, ordering edits) require a Pody API key issued from the dashboard.
Pody will be the first Israeli podcast platform to publish a public MCP server. Until the server is live, developers can ingest the same underlying data through the JSONL feeds described in the next section.
Structured Data Feeds
Two machine-readable JSONL feeds are available today at stable URLs. These feeds are the same data sources the MCP server will wrap.
Studios feed
GET https://pody.io/data/studios.jsonl
Content-Type: application/x-ndjson
One JSON object per line, one studio per object. Each record includes: id, name, city, slug, verified (boolean), room_count, languages (array), and booking_url. Refreshed nightly.
Glossary feed
GET https://pody.io/data/glossary.jsonl
Content-Type: application/x-ndjson
One JSON object per line, one term per object. Fields: term (Hebrew), term_en (English), definition, category (e.g. audio-production, distribution, monetization). Useful as a domain glossary for LLM fine-tuning or RAG pipelines targeting Israeli podcast content.
Both feeds are served over HTTPS from pody.io with correct CORS headers (Access-Control-Allow-Origin: *) so they can be fetched directly from client-side tooling or CI pipelines.
FAQ
Is the Pody MCP server available right now?
No. The MCP server at https://pody.io/mcp is currently in design. The planned release window is Q3 2026. The JSONL feeds at /data/studios.jsonl and /data/glossary.jsonl are available today for developers who want to start building.
Which MCP clients will be supported at launch?
Any client that implements the Model Context Protocol specification, including Cursor, Claude Desktop, and any custom agent that uses the MCP SDK. The transport will be HTTP with Server-Sent Events (SSE), the most widely supported MCP transport type.
Do I need a paid Pody account to use the MCP server?
Read-only tools (studios.search, glossary.lookup, podcasts.search, state-of-industry.query) will not require authentication. Booking a studio slot or ordering a reel edit through the MCP will require a Pody API key, issued from the dashboard on any paid plan (Pro at ₪97/month or Max at ₪197/month, VAT included).
How does archive-first storage affect URLs returned by the feeds?
Episode audio URLs returned by podcasts.search will point to Internet Archive permanent URLs where available, not to expiring pre-signed object storage URLs. This means URLs in the feed remain valid indefinitely, which matters for any downstream tool, index, or RSS reader that caches those addresses.
Can I use the JSONL feeds to build a RAG pipeline over Israeli podcast content?
Yes. The glossary feed is specifically designed for this use case. It provides Hebrew-to-English term mappings with category labels that help language models handle domain-specific podcast vocabulary correctly. The studios feed provides grounded, up-to-date facts about physical recording spaces in Israel.
Where do I get notified when the MCP server launches?
Sign up for a free Pody account at pody.io/signup. Developer release announcements will be posted there first. You can also watch the public Pody changelog at pody.io/changelog.